Kit ligand has a critical role in mouse yolk sac and aorta-gonad-mesonephros hematopoiesis

Abstract

Few studies report on the in vivo requirement for hematopoietic niche factors in the mammalian embryo. Here, we comprehensively analyze the requirement for Kit ligand (Kitl) in the yolk sac and aorta-gonad-mesonephros (AGM) niche. In-depth analysis of loss-of-function and transgenic reporter mouse models show that Kitl-deficient embryos harbor decreased numbers of yolk sac erythro-myeloid progenitor (EMP) cells, resulting from a proliferation defect following their initial emergence. This EMP defect causes a dramatic decrease in fetal liver erythroid cells prior to the onset of hematopoietic stem cell (HSC)-derived erythropoiesis, and a reduction in tissue-resident macrophages. Pre-HSCs in the AGM require Kitl for survival and maturation, but not proliferation. Although Kitl is expressed widely in all embryonic hematopoietic niches, conditional deletion in endothelial cells recapitulates germline loss-of-function phenotypes in AGM and yolk sac, with phenotypic HSCs but not EMPs remaining dependent on endothelial Kitl upon migration to the fetal liver. In conclusion, our data establish Kitl as a critical regulator in the in vivoAGM and yolk sac endothelial niche.

Keywords: AGM; Kit ligand; embryo; hematopoiesis; niche.

Figure 1. YS Sl/Sl EMPs fail to reach normal numbers after their initial generation

1. Number of EMPs per embryo equivalent (e.e.) of wild type and Sl/Sl E9.5 and E11.5 YS, determined by flow cytometry (panels in B,D). E9.5 Kit⁺ CD41⁺ CD16/32⁺ EMP numbers are the mean ± SD from four wild type and three Sl/Sl biological replicates, with each replicate consisting of single or two pooled YS of the same genotype. Total number of embryos analyzed: 7 +/+ (14–23 sp), 5 Sl/Sl (17–25 sp). E11.5 Kit⁺ CD41⁺ EMP numbers are the mean ± SD of 5 +/+ and 6 Sl/Sl YS analyzed individually over two independent experiments. Total live cells per YS: 1.7 ± 0.3 × 10⁵ (wt), 1.7 ± 0.2 × 10⁵ (Sl/Sl) at E9.5; 4.2 ± 1.4 × 10⁵ (wt), 4.0 ± 2.2 × 10⁵ (Sl/Sl) at E11.5.

2. Example of flow cytometry analysis of E9.5 YS EMPs quantified in panel (A). Analyses were performed on Ter119⁻ live cells.

3. Number of CFU‐C in wild type and Sl/Sl E9.5 and E11.5 YS. E9.5 data (mean ± SD) are from three wild type and two Sl/Sl biological replicates plated in duplicate, with each replicate consisting of single or two pooled YS of the same genotype. Total number of analyzed embryos: 5 +/+ (17–23 sp), 3 Sl/Sl (17–25 sp) over two independent experiments. For E11.5, data are from four wild type and six Sl/Sl biological replicates plated in duplicate, with each replicate consisting of single or two pooled YSs of the same genotype. Total number of embryos analyzed: 6 +/+, 7 Sl/Sl over four independent experiments. GEMM: granulocyte, erythroid, monocyte/macrophage, megakaryocyte; G/M/GM: granulocyte, monocyte/macrophage; Ery: erythroid.

4. Example of flow cytometry analysis of E11.5 YS EMPs quantified in panel (A). Analyses were performed on Ter119⁻ live cells.

5. Confocal WM‐IF analysis of CD31⁺ pHH3⁺ Kit⁺ proliferating EMPs (arrowheads) in E9.5 wild type and Sl/Sl YS. A single 2.5‐μm‐thick Z section is shown. For each YS, CD31⁺ Kit⁺ pHH3⁺ cells were counted in 3 to 5 fields using a 25× objective and the percentage of pHH3⁺ proliferating EMPs was calculated. Bar graphs represent the mean ± SD from three wild type (21–24 sp) and three Sl/Sl (22–23 sp) YS. Scale bars: 50 μm.

6. Cell cycle analysis of wild type and Sl/Sl E11.5 YS EMPs. The percentage of EMPs (Ter119⁻ Kit⁺ CD41⁺) in G0/G1 or S/G2/M was determined by flow cytometry on the basis of BrdU and 7‐AAD incorporation. a: apoptotic (percentage not shown). Embryos were analyzed individually. N = 6 (+/+), N = 5 (Sl/Sl) over three independent experiments. E11.5 embryos were staged according to tail somite counts as described in ref. 76 and were as follows: 12–17 (+/+); 12–17 (Sl/Sl). Bar graphs show mean ± SD, and FACS plots representative results.

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure EV1. Kitl is not required for primitive erythropoiesis or YS EMP survival

1. Number of CFU‐C in wild type and Sl/Sl E8.5 concepti (the head was removed for genotyping). Data (mean ± SD) are from three wild type and four Sl/Sl biological replicates (pools of embryos of the same genotype) plated in duplicate. Total number of analyzed embryos: 9 +/+ (3–7 sp), 9 Sl/Sl (2–8 sp) over four independent experiments.

2. Left: total number of peripheral blood (PB) cells of wild type and Sl/Sl E9.75‐E10 embryos (28–30 sp, +/+; 27–31, Sl/Sl) embryos. N = 2 (wild type), N = 5 (Sl/Sl). Right: total number of PB cells of E11.5 (tail somite range 10–16 (+/+); 12–15 (Sl/Sl)) embryos. N = 6 (wild type), N = 4 (Sl/Sl). Error bars represent SD.

3. Representative images of Wright‐Giemsa stained cytospins from E9.75‐E10 (28–30 sp, +/+; 27–31, Sl/Sl) wild type and Sl/Sl PB. Scale bars: 10 μm. N = 2 (wild type), N = 5 (Sl/Sl).

4. Flow cytometry analysis showing the percentage of early apoptotic Annexin‐V⁺ EMPs in wild type and Sl/Sl E11.5 YS. EMPs were gated as 7‐AAD⁻ Ter119⁻ CD41⁺ Kit⁺ CD16/32⁺. Embryos were analyzed in pools of 1–3. N = 9 (+/+), N = 7 (Sl/Sl) biological replicates over three independent experiments. Bar diagrams show mean ± SD values, and FACS plots representative results. Total number of analyzed embryos: 13 (+/+), 9 (Sl/Sl). Tail somite range: 11–17 (+/+); 11–17 (Sl/Sl).

5. Representative dot plots showing the flow cytometric identification of EMPs (Ter119⁻ Kit⁺ CD41⁺ CD16/32⁺) in peripheral blood (PB) of wild type and Sl/Sl midgestation embryos.

6. Percentages (mean ± SD) of EMPs in the PB of wild type and Sl/Sl embryos as identified in (E). For E9.75‐E10, PB from two wild type (28–30 sp) and five Sl/Sl (27–31) individual embryos was analyzed. For E11.5, PB from 1 to 3 embryos was pooled and six wild type and four Sl/Sl biological replicates were analyzed, with 10 wild type and 6 Sl/Sl embryos analyzed in total. *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure 2. Kitl is required for the early FL expansion of EMPs and their progeny

1. A–C

   Flow cytometry analysis of EMPs (Ter119⁻ Kit⁺ CD41⁺ CD16/32⁺) and early erythroid progenitors (Early E: Ter119⁻ Kit⁺ CD41⁻ CD71⁺ CD44⁺) in wild type and Sl/Sl E11.5 FL. EMP and early E data are the mean (± SD) of six wild type and four Sl/Sl biological replicates, with a total of 10 wild type and 6 Sl/Sl embryos analyzed in pools of 1–3 FL per genotype. Bar graphs on the right show corresponding absolute cell numbers (nr) per FL. Total FL cell counts are the mean (± SD) of 12 wild type and 13 Sl/Sl embryos. Tail somite range: 10–16 (wild type), 12–15 (Sl/Sl). Note that although EMP frequency in Sl/Sl FL is increased, their absolute number is severely decreased.

2. D

   CFU‐C numbers in wild type and Sl/Sl E11.5 FL. Data are the mean (± SD) of 10 wild type and 10 Sl/Sl biological replicates plated in duplicate, with each replicate consisting of cell from 1 to 2 FLs. A total of 12 wild type and 11 Sl/Sl FL were plated over six independent experiments. GEMM: granulocyte, erythroid, monocyte/macrophage, megakaryocyte; G/M/GM: granulocyte, monocyte/macrophage; Ery: erythroid.

3. E

   Left: Total cell numbers (nr) of E12.5 wild type (n = 7) and Sl/Sl (n = 8) FL. Right: CFU‐E numbers per wild type and Sl/Sl E12.5 FL. Data are the mean (± SD) of three biological replicates analyzed over two independent experiments. Each biological replicate consisted of 1–2 FL samples pooled according to genotype and was plated in duplicate into a methylcellulose based medium containing EPO; colonies were scored after 2 days. Total number of FLs analyzed: four wild type and six Sl/Sl.

4. F

   Quantification of erythroid lineage cells in E12.5 wild type and Sl/Sl FL. Absolute cell counts (top) and percentages (bottom) of proerythroblasts (ProE), basophilic erythroblasts (BasoE), polychromatic erythroblasts (PolyE), and orthochromatic erythroblasts (OrthoE) were determined by imaging flow cytometry analysis. The number (nr) of cells per FL was based on total cellularity and their relative proportions. Data are the mean (± SD) of three biological replicates for each genotype, with replicates consisting of single or two pooled FL samples. A total of four wild type, five Sl/+, and six Sl/Sl FLs were analyzed.

5. G

   Representative examples of maturing erythroblast series in wild type and Sl/Sl E12.5 FL, as identified by imaging flow cytometry. Scale bars: 10 μm.

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure 3. In the absence of Kitl, the influx and proliferation of EMPs in the early FL are reduced

1. A

   Number of EMPs in peripheral blood (PB) of wild type and Sl/Sl midgestation embryos, calculated from the percentage of PB EMPs (Ter119⁻ Kit⁺ CD41⁺ CD16/32⁺) determined by flow cytometry (Fig EV1E and F) and the total cell count (4.2 ± 1.6 × 10⁵ (wt), 3.1 ± 1.1 × 10⁵ (Sl/Sl) at E9.75; 2.6 ± 0.8 × 10⁶ (wt), 2.3 ± 1.4 × 10⁶ (Sl/Sl) at E11.5). For E9.75‐E10, PB from two wild type (28–30 sp) and five Sl/Sl (27–31) individual embryos was analyzed. For E11.5, PB from 1 to 3 embryos was pooled and six wild type and four Sl/Sl biological replicates were analyzed, with 10 wild type and six Sl/Sl embryos analyzed in total. Graphs show the mean (± SD) of PB EMPs per embryo equivalent (e.e.).

2. B

   Representative confocal whole‐mount immunofluorescence images of wild type and Sl/Sl FL at the onset of hematopoietic colonization. Kit⁺ EMPs are shown in white, CD31⁺ vasculature in green. Maximum intensity 3D projections from 350‐μm‐thick Z‐stacks are shown. Scale bars: 100 μm. Number of embryos analyzed: three wild type and three Sl/Sl for E9.75 (27–29 sp), seven wild type, and eight (Sl/Sl) for E10.5 (34–36 sp).

3. C

   Confocal whole‐mount immunofluorescence analysis of proliferating hematopoietic progenitors in wild type and Sl/Sl E10.5 (34–36 sp) FL. Very few CD31⁺ pHH3⁺ Kit⁺ proliferating hematopoietic cells (arrowheads) were detected in the Sl/Sl FL. A single 2.5‐μm‐thick slice is shown. Scale bars: 20 μm. A total of seven wild type and eight Sl/Sl embryos were analyzed.

4. D, E

   Percentage of cycling Ter119⁻ Kit⁺ CD41⁺ CD16/32⁺ EMPs and Ter119⁻ Kit⁺ CD41⁻ CD71⁺ CD44⁺ early erythroid (Early E) cells in the E11.5 FL, determined by flow cytometric analysis of BrdU and 7‐AAD incorporation. Data are the mean (± SD) of six wild type and five Sl/Sl FLs analyzed individually over three independent experiments. Tail somite range: 12–17 (+/+); 12–17 (Sl/Sl).

5. F, G

   Percentage of early apoptotic EMPs and Early E in wild type and Sl/Sl E11.5 FL, determined by flow cytometric analysis of Annexin‐V staining among 7‐AAD⁻ Ter119⁻ CD41⁺ Kit⁺ CD16/32⁺ EMPs and 7‐AAD⁻ Ter119⁻ Kit⁺ CD41⁻ CD71⁺ CD44⁺ Early E. Data are the mean (± SD) of nine wild type and seven Sl/Sl biological replicates analyzed over three independent experiments, with each replicate consisting of 1‐3 FLs of identical genotypes. A total number of 13 wild type and 9 Sl/Sl embryos were analyzed. Tail somite range: 11–17 for both genotypes.

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure 4. Macrophages are depleted in Sl/Sl mutant embryos

1. Percentages of macrophage populations in wild type and Sl/Sl E14.5 brain, skin (dissected from the back), lungs, and limb buds, as determined by flow cytometry. Cells were gated as Ter119⁻ CD45⁺ and further gated into F4/80^hi CD11b^lo or F4/80^lo CD11b^hi populations. Representative dot plots of skin analysis are shown in Appendix Fig S1A. Data are the mean (± SD) of four biological replicates for either genotype, with each replicate consisting of individual or pooled samples (from up to three embryos). A total of eight wild type and five Sl/Sl embryos were analyzed over two independent experiments.

2. Percentage and absolute numbers of tissue macrophages per E14.5 FL. Data are mean (± SD). Number of embryos analyzed: as in (A).

3. Representative flow cytometry analysis of macrophages in wild type and Sl/Sl E10.5 AGM+VU. Cells were gated as shown, with percentages in individual gates as indicated. Data are the mean (± SD) of four wild type (32–35 sp) or three Sl/Sl (34–35 sp) biological replicates, analyzed over two independent experiments. Embryos were analyzed individually.

4. Quantification of flow cytometry analysis in (C), showing percentage of F4/80⁺CD11b⁺ macrophages within live cells. E11.5 data are the mean (± SD) of four wild type (10–13 tail somites) or three Sl/Sl (10–13 tail somites) biological replicates, analyzed over three independent experiments. Embryos were analyzed individually.

5. Confocal whole‐mount immunofluorescence analysis of macrophages in wild type and Sl/Sl E10.5 AGM+VU. 3D panels show maximum intensity projections from 400‐μm‐thick Z‐stacks. Slice panels show single 2.5‐μm‐thick Z‐slices of clusters in the dorsal aorta (da). A total of four wild type (31–34 sp) and three Sl/Sl (31–33 sp) embryos were analyzed. Scale bars: 50 μm (3D), 10 μm (slice).

6. Quantification of whole‐mount immunofluorescence in (E), showing absolute numbers (mean ± SD) of F4/80⁺ macrophages within hematopoietic clusters in the dorsal aorta.

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure EV2. Lymphomyeloid immune‐restricted progenitor (LMPP) numbers are reduced in Sl/Sl FL and PB

1. Flow cytometry analysis of Lin⁻ CD19⁻ B220⁻ CD45⁺ Kit⁺ Flt3⁺ IL7Rα⁺ LMPPs in wild type and Sl/Sl E11.5 FL. Lineage cocktail (Lin): Ter119, CD3e, F4/80, Nk1.1, Gr1. Gates and percentages within gate are indicated in representative dot plots. Percentages are the mean (± SD) of six wild type and four Sl/Sl biological replicates, with each replicate consisting of FL cells from individual or pools of up to three embryos. A total of 10 wild type (10–16 tail sp) and 6 Sl/Sl (12–15 tail sp) embryos were analyzed.

2. Percentage and number of LMPP per E11.5 FL, as identified by flow cytometry and total FL cell counts. Data are the mean (± SD); samples as in (A).

3. Percentage and number of LMPP in PB, as identified by flow cytometry and PB cell counts per embryo. Data are the mean (± SD); samples as in (A). e.e.: embryo equivalent.

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure 5. Kitl is required for AGM (pre‐)HSC maturation and survival in vivo

1. Confocal whole‐mount immunofluorescence analysis (WM‐IF) of Kit⁺ cells in wild type and Sl/Sl E10.5 AGM (32–36 sp). Images are 3D projections from 400‐μm‐thick Z‐stacks, and are representative of 14 wild type and 11 Sl/Sl embryos analyzed. Total numbers of Kit⁺ cells (white) in the dorsal aorta were quantified and showed a significant reduction in Sl/Sl compared to wild type (P = 0.008). CD31⁺ vasculature is shown in green. Arrowheads: hematopoietic cell clusters. Scale bars: 100 μm.

2. Quantification of WM‐IF in (A), showing absolute numbers (± SD) of clusters of indicated size and separately shown relative to location. A total of seven wild type and seven Sl/Sl embryos were counted.

3. Flow cytometry analysis of the HSC lineage in wild type and Sl/Sl E11.5 AGM+VU. Cells were Ter119⁻ VE‐Cadherin⁺ Kit⁺ gated, and percentages of cells within the pre‐HSC type I gate (red) and the pre‐HSC type II/HSC gate (green) are shown. Percentages are the mean (± SD) of 10 wild type and 9 Sl/Sl biological replicates, with each replicate consisting of 1–3 pooled AGM+VU. A total of 15 wild type and 11 Sl/Sl embryos were analyzed over five independent experiments.

4. Absolute numbers (± SD) of E10.5 and E11.5 pre‐HSC type I (Ter119⁻ VE‐Cadherin⁺ Kit⁺ CD41⁺ CD45⁻) and pre‐HSC type II/HSC (Ter119⁻ VE‐Cadherin⁺ Kit⁺ CD41⁺ CD45⁺), as determined by flow cytometry in (C). Number of E11.5 embryos analyzed: as in (A). For E10.5, five biological replicates were analyzed for either genotype, with each replicate consisting of 1–4 pooled AGM+VU. A total of 12 wild type (31–35 sp) and 9 Sl/Sl (30–34 sp) embryos were analyzed over four independent experiments. Total live cells per AGM+VU: 2.8 ± 1.0 × 10⁵ (wt), 2.3 ± 0.6 × 10⁵ (Sl/Sl) at E10.5; 3.9 ± 1.1 × 10⁵ (wt), 3.8 ± 1.4 × 10⁵ (Sl/Sl) at E11.5.

5. CFU‐C numbers per wild type and Sl/Sl E11.5 AGM+VU. Counts are the mean (± SD) of four wild type and six Sl/Sl biological replicates plated in duplicate. Biological replicates consisted of cells from individual or two pooled AGM+VU. A total of six wild type and seven Sl/Sl AGM+VU embryos were analyzed over four independent experiments. GEMM: granulocyte, erythroid, monocyte/macrophage, megakaryocyte; G/M/GM: granulocyte, monocyte/macrophage; Ery: erythroid.

6. Analysis of long‐term multi‐lineage HSC potential in E11.5 Sl/Sl AGM+VU. Irradiated CD45.1 syngeneic mice were transplanted with 1 e.e. of wild type, Sl/+, or Sl/Sl CD45.2⁺ AGM+VU cells. PB chimerism is represented as the percentage of donor CD45.2⁺ cells among total CD45⁺ cells, 16 weeks after transplant. A total of 14 recipients were transplanted with wild type or Sl/+ cells and 8 with Sl/Sl cells, over four independent experiments. Wild type is represented with a circle, Sl/+ a triangle, and Sl/Sl with a square. Tail somite range: 9–17 (+/+, Sl/+); 10–17 (Sl/Sl). *P = 0.027, one‐tailed Mann–Whitney U‐test; P = 0.055, two‐tailed Mann–Whitney U‐test.

7. Confocal WM‐IF analysis of pHH3 expression in Kit⁺ cells of wild type and Sl/Sl E10.5 dorsal aorta (33–36 sp). Images are 2.5‐μm‐thick single longitudinal Z‐slices; dorsal up. Arrowheads indicate examples of proliferating CD31⁺ pHH3⁺ Kit⁺ hematopoietic cells (magnified in insets); the total number of these cells per dorsal aorta was quantified and showed reduced proliferation in the mutant (P = 0.01). A total of four embryos per genotype were analyzed. Scale bars: 50μm.

8. Confocal WM‐IF analysis of cCasp3 expression in Kit⁺ cells of wild type and Sl/Sl E10.5 dorsal aorta (33–36 sp). 2.5‐μm‐thick single longitudinal Z‐slices; dorsal up. A higher number of Kit⁺ cCasp3⁺ apoptotic cluster cells were detected in the Sl/Sl dorsal aorta (P = 0.04). Arrowheads indicate a CD31⁺ Kit⁺ cCasp3⁻ hematopoietic cluster for wild type and an apoptotic CD31⁺ Kit⁺ cCasp3⁺ hematopoietic cell for Sl/Sl (both magnified in insets). A total of four wild type and three Sl/Sl embryos were analyzed. Scale bars: 50 μm.

9. Analysis of apoptosis among pre‐HSC type II/HSCs in wild type and Sl/Sl E11.5 AGM. Ter119⁻ VE‐Cadherin⁺ Kit⁺ CD41⁺ CD45⁺ cells were analyzed by flow cytometry for Annexin‐V binding to detect early apoptotic cells. Percentages are the mean (± SD) of nine wild type and seven Sl/Sl biological replicates analyzed over three independent experiments. Individual replicates consisted of 1–3 AGMs. A total of 13 wild type and 9 Sl/Sl embryos (11–17 tail sp) were analyzed. The percentage of apoptotic pre‐HSC type II/HSCs in the mutant AGM was significantly increased (P = 0.04).

10. Multiplex Fluidigm qRT–PCR analysis of pre‐HSC type I and II, sorted as shown in (C) from E11.5 wild type (n = 4 biological replicates; total number of embryos: 9; 11–17 tail somites) and Sl/Sl (n = 5 biological replicates; total number of embryos: 9; 11–17 tail somites). Data are mean (± SD).

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test in all panels except F).

Figure EV3. Transcriptional analysis of wild type and Sl/Sl AGM hematopoietic clusters and endothelium

1. E11.5 AGM+VU Ter119⁻ VE‐Cadherin⁺ Kit⁺ cells (hematopoietic clusters) were sorted from wild type and Sl/Sl embryos (pools of 1–4 embryos of the same genotype; 2 sorting experiments). Duplicate samples of 100 cells each were processed using SMARTSEQ2 80. The un‐clustered heat map shows 228 genes differentially expressed (P < 0.05, FDR < 0.1) between wild type and Sl/Sl hematopoietic clusters as detected by RNA‐Seq and is ordered by log(fold change) with high positive values on top and negative values on bottom. Each column shows a biological replicate.

2. Gene ontology (GO) analysis performed with the MetaCore suite (Thomson Reuters) of the 213 upregulated and 15 downregulated genes identified by RNA‐Seq, showing selected categories within the top over‐represented GO processes.

3. E11.5 AGM+VU Ter119⁻ VE‐Cadherin⁺ Kit⁻ CD41⁻ CD45⁻ (endothelial cells) were sorted from wild type and Sl/Sl embryos (pools of 1–4 embryos of the same genotype; 2 sorting experiments). Duplicate samples of 100 cells each were processed using SMARTSEQ2. The un‐clustered heat map shows 70 genes differentially expressed (P < 0.05, FDR < 0.1) between wild type and Sl/Sl endothelium as detected by RNA‐Seq and is ordered by log(fold change) with high positive values on top and negative values on bottom. Each column shows a biological replicate.

4. Gene ontology (GO) analysis performed with the MetaCore suite (Thomson Reuters) of the 33 upregulated and 37 downregulated genes identified by RNA‐Seq, showing selected categories within the top over‐represented GO processes.

5. RPKM expression of genes encoding for phenotypic markers used to identify and sort cell populations, in wild type and Sl/Sl E11.5 AGM+VU hematopoietic clusters (Ter119⁻ VE‐Cadherin⁺ Kit⁺) and endothelium (Ter119⁻ VE‐Cadherin⁺ Kit⁻ CD41⁻ CD45⁻). Note low expression of Kit in the endothelium, and absence of Kitl in Sl/Sl (as expected). *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure EV4. Validation and quantification of Kitl‐tdTomato reporter expression

1. Kitl‐tdTomato⁺ and Kitl‐tdTomato⁻ subsets were isolated from pooled E8.5 (3–8 sp) Kitl‐tdTomato transgenic concepti and E10.5 (31–36 sp) AGM regions (including vitelline and umbilical arteries) (AGM+VU), YS, and FL. PAS: para‐aortic splanchnopleura. Sort gates as indicated. Total number of Kitl‐tdTomato transgenic embryos analyzed: 14 (E8.5), 21 (E10.5). N = 2 independent experiments for E8.5; N = 2 independent experiments for E10.5.

2. qRT–PCR analysis of endogenous Kitl in the Kitl‐tdTomato⁺ and Kitl‐tdTomato⁻ subsets shown in (A). Endogenous Kitl expression (relative to Atp5a1) is virtually restricted to Kitl‐tdTomato⁺ cells.

3. Control sample for confocal whole‐mount immunofluorescence analysis of Kitl‐tdTomato expression in the E8.5 (6–7 sp) YS BI. The full staining (using CD31, GFP and anti‐RFP/tdTomato antibodies) has been performed on a 23GFP embryo. No significant background is observed. A single 2.5‐μm‐thick Z‐slice is shown. Scale bar: 50 μm.

4. Flow cytometry analysis of E10.5 (32–37 sp) Kitl‐tdTomato transgenic YS. Live cells were gated as total Ter119⁻, or as Ter119⁻ endothelial cells (EC: VE‐Cadherin⁺ Kit⁻ CD41⁻ CD45⁻), hematopoietic clusters (HC: VE‐Cadherin⁺ Kit⁺), hematopoietic progenitors (HP: VE‐Cadherin⁻ Kit⁺), and mesenchymal cells (MC: VE‐Cadherin⁻ Kit⁻ CD41⁻ CD45⁻). Representative dot plots are shown. Nineteen Kitl‐tdTomato transgenic embryos were analyzed (pooled tissues) in a total of three independent experiments. Data in the graph are mean (± SD).

5. Confocal whole‐mount immunofluorescence analysis of a E9.5 (25–26 sp) Kitl‐tdTomato embryo. Left image shows a maximum intensity 3D projection from a 500‐μm‐thick Z‐stack. The region indicated in the boxed inset is magnified in the middle and right panels, which show a single 2.5‐μm‐thick longitudinal slice. Arrowheads indicate Kitl‐tdTomato⁺ CD31⁺ endothelial cells in the ventral wall of the dorsal aorta. Scale bars: 150 μm (3D projection); 50μm (slice inset). va: vitelline artery; da: dorsal aorta. N = 3 embryos analyzed.

6. Flow cytometry analysis of E10.5 Kitl‐tdTomato transgenic AGM+VU from the same embryos as in (D). Representative dot plots are shown. Live cells were gated as Ter119⁻ or as Ter119⁻ EC, HC, MC, as shown. Data in the graph are mean (± SD).

7. Flow cytometry analysis of Kitl‐tdTomato expression in the non‐hematopoietic (Ter119⁻ CD41⁻ CD45⁻) components of the E10.5 (32–37 sp) FL. Endothelial (EC: VE‐Cadherin⁺ Dlk1⁻), mesenchymal (MC: VE‐Cadherin⁻ Dlk1⁻) cells, and hepatoblasts (HB: VE‐Cadherin⁻ Dlk1⁺) were gated as indicated. Representative dot plots with percentage of Kitl‐TdTomato⁺ cells are shown. HB make up the majority of the non‐hematopoietic FL cells and, in line with previous reports 65, 66, show extensive Kitl expression. A total of 19 Kitl‐tdTomato embryos, pooled according to genotype, were analyzed over three independent experiments (N = 2 experiments for MC and HB. Kitl‐tdTomato⁺ range: MC 23.1–40.4%; HB 84.1–86.9%). Data in the graph are mean (± SD).

Figure 6. Spatiotemporal expression of Kitl‐tdTomato at pre‐liver sites of hematopoiesis

1. Confocal whole‐mount immunofluorescence (WM‐IF) analysis of Kitl‐tdTomato expression in the E8.5 (6–7 sp) conceptus (maximum intensity 3D projection from a 500‐μm‐thick Z‐stack). Kitl‐tdTomato expression (red) is seen in the YS, allantois (a; from which the umbilical artery derives), and para‐aortic splanchnopleura (PAS) of Kitl‐tdTomato::23GFP transgenic concepti (the embryo proper, from which the head has been removed, is outlined with a purple dashed line). The 23GFP transgene (green) labels the YS blood islands (bi; yellow dashed line); this GFP reporter is transcribed from a heterologous hsp68 promoter under the spatiotemporal control of the Runx1 hematopoietic +23 enhancer, resulting in GFP expression in all primitive erythroid(‐fated) cells, hemogenic endothelium, and emerging HSPCs 2, 74. Endothelial cells (CD31⁺) are in white. Arrowheads indicate the paired aortae (pa). Purple boxed area is magnified in (B); yellow boxed area is magnified in (C). Scale bar: 100 μm. N = 2 embryos analyzed.

2. Single 2.5‐μm‐thick Z‐slice from the region highlighted in the purple box in (A). Arrowheads indicate Kitl‐tdTomato⁺ CD31⁺ endothelial cells in one of the paired aortae (pa). Scale bar: 25 μm.

3. Single 2.5‐μm‐thick Z‐slice from the region highlighted in the yellow box in (A), showing Kitl‐tdTomato expression in the E8.5 YS BI. Arrowheads indicate Kitl‐tdTomato⁺ CD31⁺ endothelial cells; the arrow indicates a Kitl‐tdTomato⁺ perivascular cell. Scale bar: 25 μm.

4. Single 2.5‐μm‐thick Z‐slice of 10.5 (35 sp) YS (WM‐IF) showing Kitl‐tdTomato (red), Kit (white), and CD31 (green) expression. Arrowheads indicate Kitl‐tdTomato⁺ CD31⁺ endothelial cells. N = 3 embryos analyzed. Scale bar: 50 μm.

5. Confocal WM‐IF analysis of Kitl‐tdTomato expression (red) in a E10.5 (36 sp) AGM region. Top panels (3D view) show a maximum intensity 3D projection from a 400‐μm‐thick Z‐stack. Bottom panels (slice view) show a single 2.5‐μm‐thick Z‐slice of the dorsal aorta, from the boxed region in the top panels. CD31 (vasculature; green); Kit (hematopoietic progenitors; white). Scale bars: 100 μm (top); 25 μm (bottom). fl: fetal liver; va: vitelline artery plexus; da: dorsal aorta; fp: floor plate; nt: neural tube. N = 6 embryos analyzed.

6. Confocal immunofluorescence analysis of cryosections through the AGM region of E10.5 (34–36 sp) Kitl‐tdTomato transgenic embryos. Arrowheads indicate Kitl‐tdTomato⁺ CD31⁺ endothelial cells (top), Kitl‐tdTomato⁺ PDGFR‐β⁺ (middle), or Kitl‐tdTomato⁺ α‐SMA⁺ (bottom) perivascular smooth muscle cells. Scale bars: 25 μm or 50 μm (PDGFR‐β panel). N > 6 embryos analyzed.

7. Confocal immunofluorescence analysis of Kitl‐tdTomato expression in the E10.5 FL. A cryosection through a E10.5 (35–36 sp) Kitl‐tdTomato transgenic FL is shown. Arrowheads indicate Kitl‐tdTomato⁺ CD31⁺ Kit⁻ endothelial cells. Asterisks indicate Kitl‐tdTomato⁺ CD31⁻ Kit⁻ mesenchymal cells or hepatoblasts. Note proximity of Kitl‐tdTomato⁺ cells and Kit⁺ hematopoietic cells. Scale bars: 50 μm (left panel), 25 μm (magnified inset). N > 5 embryos analyzed.

Figure 7. Endothelial Kitl is required for hematopoietic stem and progenitor cells in the AGM and for EMPs in the YS

1. Schematic showing the two genetically modified mouse lines crossed to perform conditional deletion of Kitl in endothelial cells.

2. Left: number of Ter119⁻ Kit⁺ CD41⁺ CD16/32⁺ EMPs in Tie2Cre::Kitl ^Δ/Δ and control Kitl ^{f/f or f/+} E11.5 YS, determined by flow cytometry. Data are the mean (± SD) of eight control and seven Tie2Cre::Kitl ^Δ/Δ YS analyzed individually over four independent experiments. Tail somite range: 13–16. Middle: number of pre‐HSC type I (Ter119⁻ VE‐Cadherin⁺ Kit⁺ CD41⁺ CD45⁻) and pre‐HSC type II/HSC (Ter119⁻ VE‐Cadherin⁺ Kit⁺ CD41⁺ CD45⁺) in Tie2Cre::Kitl ^Δ/Δ and control Kitl ^{f/f or f/+} E11.5 AGM+VU, determined by flow cytometry. Data are the mean (± SD) of seven control and five Tie2Cre::Kitl ^Δ/Δ AGM+VU analyzed individually over three independent experiments. Tail somite range: 14–15 (Kitl ^{f/f, f/+}); 14–16 (Tie2Cre::Kitl ^Δ/Δ). Right: number of Ter119⁻ Kit⁺ CD41⁺ CD16/32⁺ EMPs and Ter119⁻ Kit⁺ CD41⁻ CD71⁺ CD44⁺ early erythroid cells in Tie2Cre::Kitl ^Δ/Δ and control Kitl ^{f/f or f/+} E11.5 FL, determined by flow cytometry. Data are the mean (± SD) of seven control and five Tie2Cre::Kitl ^Δ/Δ FL analyzed individually over three independent experiments. Tail somite range: 13–16.

3. CFU‐C numbers in Tie2Cre::Kitl ^Δ/Δ and control Kitl ^{f/f or f/+} E11.5 YS. Data are the mean (± SD) of six control and three Tie2Cre::Kitl ^Δ/Δ YS plated in duplicate in two independent experiments. Tail somite range: 13–16.

4. Analysis of long‐term multi‐lineage HSC potential in E11.5 Tie2Cre::Kitl ^Δ/Δ and control Tie2Cre::Kitl ^Δ/+ or Kitl ^{f/f or f/+}AGM+VU. Irradiated CD45.1 syngeneic mice were transplanted with 1 e.e. of AGM+VU cells. PB chimerism is represented as the percentage of donor CD45.2⁺ cells among total CD45⁺ cells, 16 weeks after transplant. A total of 29 recipients were transplanted with Tie2Cre::Kitl ^Δ/+ or Kitl ^{f/f or f/+} cells and eight with Tie2Cre::Kitl ^Δ/Δ cells, over seven independent experiments. Kitl ^{f/f or f/+} is represented with a circle, Tie2Cre::Kitl ^Δ/+ a triangle and Tie2Cre::Kitl ^Δ/Δ with a square. Tail somite range: 12–17 (control); 12–17 (Tie2Cre::Kitl ^Δ/Δ).

5. CFU‐C numbers in Tie2Cre::Kitl ^Δ/Δ and control Kitl ^{f/f or f/+} E11.5 Fl. Data are the mean (± SD) of 7 control and 5 Tie2Cre::Kitl ^Δ/Δ FL plated in duplicate in three independent experiments. GEMM: granulocyte, erythroid, monocyte/macrophage, megakaryocyte; G/M/GM: granulocyte, monocyte/macrophage; Ery: erythroid. Tail somite range: 13–16.

6. Number of phenotypic LSK, HPC, and HSC per FL in E12.5 Tie2Cre::Kitl ^Δ/Δand control (Kitl ^{f/f or f/+}) FL, determined by flow cytometry and total FL cell counts. Cells were gated as 7‐AAD⁻ Lin⁻ (F4/80⁻ CD3e⁻ Nk1.1⁻ Ter119⁻ Gr‐1⁻ B220⁻ CD19⁻). LSK: Lin⁻ Sca1⁺ Kit⁺; HPC: Lin⁻ Sca1⁺ Kit⁺ CD48⁺ CD150⁻ and HSC: Lin⁻ Sca1⁺ Kit⁺ CD48⁻ CD150⁺. Gating strategy is shown in Fig EV4E. Data are the mean (± SD) of 12 control and 6 Tie2Cre::Kitl ^Δ/Δ FLs analyzed individually over three independent experiments.

7. Total cellularity of E12.5 Kitl ^{f/f or f/+} (n = 14) and Tie2Cre::Kitl ^Δ/Δ (n = 11) FL. Error bars represent SD.

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure EV5. Effects of Kitl loss on AGM and FL HSCs

1. A

   Analysis of long‐term multi‐lineage HSC potential in E11.5 Tie2Cre::Kitl ^Δ/Δ and control Tie2Cre::Kitl ^Δ/+ or Kitl ^{f/f or f/+}AGM+VU. Irradiated CD45.1 syngeneic mice were transplanted with 1 e.e. of AGM+VU cells. PB chimerism is represented as the percentage of donor CD45.2⁺ cells among total CD45⁺ cells, 6 weeks after transplant. A total of 29 recipients were transplanted with Tie2Cre::Kitl ^Δ/+ or Kitl ^{f/f or f/+} cells and 8 with Tie2Cre::Kitl ^Δ/Δ cells, over seven independent experiments. Kitl ^{f/f or f/+} is represented with a circle, Tie2Cre::Kitl ^Δ/+ a triangle and Tie2Cre::Kitl ^Δ/Δ with a square. Tail somite range: 12–17 (control); 12–17 (Tie2Cre::Kitl ^Δ/Δ).

2. B

   Analysis of phenotypic‐defined HSPCs populations in wild type and Sl/Sl E12.5 FL. Representative dot plots with gates and percentages of 7‐AAD⁻ Lin⁻ (F4/80⁻ CD3e⁻ Nk1.1⁻ Ter119⁻ Gr‐1⁻ B220⁻ CD19⁻) LSK, HPC, and HSC are shown. Fetal livers were analyzed individually. N = 4 (wild type); N = 5 (Sl/Sl).

3. C

   Number of LSK, HPC, and HSC per FL, determined by flow cytometry in (E) and total FL cell counts. Number of FL analyzed at E12.5: as in (D). E14.5 data are the mean (± SD) of five biological replicates for either genotype. Replicates consisted of cells from 1 to 3 FLs, with a total of 10 wild type and 5 Sl/Sl FLs analyzed. *P < 0.05 (unpaired two‐tailed Student's t‐test).

Figure 8. Kitl is required for Akt and Erk phosphorylation in the AGM and YS

1. A, B

   Confocal whole‐mount immunofluorescence analysis of phosphorylated Akt (pAkt) (A) and phosphorylated p44/42 MAPK (pErk1/2) (B) in Kit⁺ cells of wild type (32–36 sp) and Sl/Sl (31–35 sp) E10.5 dorsal aorta, yolk sac, and fetal liver. Images are 2.5‐μm‐thick single longitudinal Z‐slices; dorsal up (DA). Arrowheads point to Kit⁺ cells with varying levels of pAkt and pErk1/2 expression. A total of two wild type and two Sl/Sl embryos for each staining were analyzed. Scale bars: 20 μm.

2. C, D

   Quantification of the immunofluorescence in (A, B), showing mean fluorescence intensity (MFI) ratio of pAkt and p44/42 MAPK (pErk1/2) within wild type (+/+) and Sl/Sl CD31⁺ Kit⁺ dorsal aorta (DA), yolk sac (YS), and fetal liver (FL) hematopoietic cells. MFI values were normalized to background in the adjacent ventral mesenchyme for DA, to erythrocytes in the YS or stroma in the FL. Additional quantification was performed on sections from two wild type and two Sl/Sl embryos of similar age range. MFI measurements were taken using ImageJ. Number of cells measured for pAkt: 100 hematopoietic cells and 73 background (+/+); 79 hematopoietic and 52 background (Sl/Sl) in the AGM; 127 hematopoietic cells and 95 background (+/+); 110 hematopoietic and 82 background (Sl/Sl) in the YS; 102 hematopoietic cells and 78 background (+/+); 78 hematopoietic and 83 background (Sl/Sl) in the FL. Number of cells measured for pErk1/2: 94 hematopoietic cells and 82 background (+/+); 114 hematopoietic and 91 background (Sl/Sl) in the AGM; 174 hematopoietic cells and 109 background (+/+); 123 hematopoietic and 82 background (Sl/Sl) in the YS; 100 hematopoietic cells and 70 background (+/+); 63 hematopoietic and 57 background (Sl/Sl) in the FL. N = 4 (+/+), 4 (Sl/Sl) embryos analyzed for DA and YS; N = 3 (+/+), 2 (Sl/Sl) for FL. Error bars represent SD.

3. E

   Multiplex Fluidigm qRT–PCR analysis of pre‐HSC type I (Ter119⁻ VE‐Cadherin⁺ Kit⁺ CD41⁺ CD45⁻), pre‐HSC type II/HSC (Ter119⁻ VE‐Cadherin⁺ Kit⁺ CD41⁺ CD45⁺), YS and FL EMP (Ter119⁻ Kit⁺ CD41⁺ CD16/32⁺), FL Early E (Ter119⁻ Kit⁺ CD41⁻) and peripheral blood LMPP (Lin⁻ CD19⁻ B220⁻ CD45⁺ Kit⁺ Flt3⁺ IL7Rα⁺), isolated from E11.5 (11–17 tail somites) wild type and Sl/Sl embryos. Four biological replicates for wild type (9 total embryos) and five biological replicates (9 total embryos) for Sl/Sl were analyzed. Data are mean (± SD).

Data information: *P < 0.05 (unpaired two‐tailed Student's t‐test).